Effects of Perioperative Antiinflammatory and Immunomodulating Therapy on Surgical Wound Healing

Anthony J. Busti, PharmD; Justin S. Hooper, PharmD; Christopher J. Amaya, PharmD; Salahuddin Kazi, MBBS


Pharmacotherapy. 2005;25(11):1566-1591. 

In This Article

Drugs Used in Rheumatoid Arthritis and Crohn's Disease and Their Effects on Wound Healing

Methotrexate. The DMARDs such as methotrexate, D-penicillamine, and azathioprine play an important role in slowing the progression of disease in patients with various autoimmune disorders. Methotrexate exhibits its cytotoxic effect by inhibiting purine nucleotide and thymidylate synthesis through the blockade of dihydrofolate reductase.[61,62] After oral administration, methotrexate is metabolized into polyglutamates, which concentrate intracellularly, thus forming a drug depot.[63,64] Renal elimination of methotrexate is dependent on its free, unbound serum concentration, thereby requiring close monitoring of patients with renal impairment who are receiving methotrexate therapy.[65] This is substantiated by the finding that the risk of overall toxicity due to methotrexate increases as renal function decreases.[66] In addition, patients taking methotrexate are at risk for hematologic toxicity irrespective of the duration of methotrexate treatment. One group of authors suggests that temporary discontinuation of methotrexate should be considered in patients with advanced age, renal impairment, sepsis, or undergoing major surgery (especially if a combination of these factors are present) because of the risk for pancytopenia.[67]

Methotrexate may affect the human production or physiologic response to inflammatory mediators such as leukotrienes, cytokines, and chemokines.[62] In six patients who received long-term methotrexate therapy at a mean dosage of 9.6 mg/week, predose leukotriene B4 levels in neutrophils were significantly higher than levels measured 1 day after dosing (13.0 vs 6.0 ng/10[6] cells, p=0.0019).[68] Another group used the murine air pouch model as an in vivo model of inflammation to demonstrate that low-dose, weekly administration of methotrexate in mice led to adenosine-dependent inhibition of inflammation.[69] The same investigators also showed that an adenosine analog, 2-chloroadenosine, prevents the adherence of stimulated neutrophils to endothelial cells.[70] In addition, methotrexate decreases polymorphonuclear leukocyte chemotaxis,[71,72] which is another necessary component of the cellular response phase of wound healing.

Postoperative complications may arise as a result of wound infection in patients taking methotrexate. According to the Arthritis, Rheumatism, and Aging Medical Information System, nonsurgical patients with rheumatoid arthritis who are taking methotrexate have an incidence of infection of approximately 17/1000 person-years compared with none with D-penicillamine and 5/1000 with azathioprine.[73] Methotrexate therapy has been associated with opportunistic infections caused by pathogens such as cytomegalovirus,[74] Nocardia asteroides,[75]Listeria monocytogenes,[76] herpes zoster,[77]Pneumocystis carinii,[78,79] and Aspergillus,[80] which may also cause delays in postoperative wound healing.

A number of clinical trials have shown little to no effect of methotrexate on recovery after orthopedic surgery ( Table 2 ).[81,82,83,84,85,86,87,88,89,90,91] Also, no significant risk of postoperative complications was found among approximately 220 patients who received low-dose pulse methotrexate across four retrospective studies.[81,82,83,84] All patients had rheumatoid arthritis and underwent surgery for reconstruction of the hand and wrist, total joint arthroplasty, or cervical spine fusion. The results of these trials are confirmed by two prospective trials in which the authors concluded that there is little justification for the withdrawal of methotrexate therapy before orthopedic surgery because of the low rate of postoperative complications.[85,86] In the first study, 388 patients taking methotrexate 2.5-25 mg/week were randomly assigned to either continue methotrexate therapy throughout surgery or withhold methotrexate from 2 weeks before to 2 weeks after surgery.[85] In the other study, 64 patients with rheumatoid arthritis who were taking methotrexate 5-15 mg/week were randomly assigned to either continue methotrexate or discontinue it 7 days before surgery.[86] Patients in these trials were also taking other drugs including D-penicillamine, indomethacin, and prednisolone. Compared with patients who held their methotrexate therapy for either 1 or 2 weeks before surgery, those who continued methotrexate therapy throughout the perioperative period in each of these two trials had no significant increase in wound infection or delayed healing.

Conversely, in another retrospective study, authors analyzed 38 patients undergoing elective orthopedic surgery.[87] Patients were divided into two groups according to the time of methotrexate discontinuation before surgery. When methotrexate was discontinued less than 4 weeks before surgery, four of 19 patients experienced early postoperative complications such as wound dehiscence and bacterial infection. Three of the four patients with complications were recipients of prosthetic knee replacements. In patients who discontinued methotrexate therapy 4 or more weeks before surgery, no postoperative complications were identified. These patients resumed normal methotrexate treatment within 4-6 weeks after surgery, but many subjects reported worsening pain and stiffness while not taking their normal regimen of methotrexate. Thus, the potential benefit of early withdrawal of methotrexate to avert postoperative complications may be outweighed by decreased quality of life and possible flare-up of disease. The authors of the study acknowledge that their small sample and retrospective design should prompt the initiation of a larger prospective study to determine if preoperative methotrexate increases the risk of postoperative complications.

Another prospective study examined the effect of methotrexate on postoperative complications in patients with rheumatoid arthritis who were undergoing total joint arthroplasty or joint fusion.[88] Patients were nonrandomly assigned to either discontinue methotrexate starting the week before surgery (group 1, 19 patients) or continue methotrexate throughout the peri-operative period with no discontinuation or decrease in methotrexate dosage (group 2, 13 patients). The mean weekly dose of methotrexate was 12.5 mg in group 1 and 13.1 mg in group 2, and the mean daily prednisone dose was 6.8 and 6.1 mg, respectively. Forty-two procedures were performed on 32 patients who were included. No postoperative infections were noted in group 1, whereas four infections occurred in group 2 within the first month after surgery (p=0.03). No patients in either group experienced a post-operative disease flare-up. The authors concluded that based on their findings it was prudent to discontinue methotrexate therapy 1 week before surgery in order to decrease the risk of infection while at the same time preventing a flare-up of rheumatoid arthritis.

In summary, numerous retrospective and prospective trials have examined the effects of methotrexate therapy on postoperative complications in patients with rheumatoid arthritis who are undergoing orthopedic surgery. Because of the nature of the disease, many of the patients studied were also taking other DMARDs, NSAIDs, or corticosteroids, which may have confounded the end points. In addition, the weak study designs and lack of power analyses performed in these trials add to the difficulty in determining the true effect. Therefore, a careful analysis of each patient and the risks and benefits of either stopping or continuing methotrexate therapy before surgery should be conducted. It appears reasonable to hold methotrexate for 2-4 weeks before surgery in patients who do not have severe, debilitating disease or who would not otherwise be incapacitated by the withdrawal of methotrexate therapy.

D-penicillamine. D-penicillamine is indicated for the treatment of Wilson's disease, cystinuria, and severe rheumatoid arthritis that fails to respond to conventional therapy (although D-penicillamine is seldom used for the latter indication). When used for the treatment of rheumatoid arthritis, it is thought to depress circulating immunoglobulin M rheumatoid factor and helper T cell activity.[92] Since D-penicillamine has antifibrotic activity,[93,94] its use may be detrimental to the normal wound healing process. Specifically, it has been reported to prevent maturation of the insoluble polymerized form of collagen,[95,96] decrease collagen and protein synthesis,[97] and chelate cross-linking sites on collagen fibers.[95,98]

Investigators studied the effects of D-penicil-lamine 600 mg/day on wound healing in patients undergoing orthopedic surgery.[89] These investigators found that only 1 of 48 patients experienced delayed wound healing and concluded that treatment with D-penicillamine should not be discontinued before surgery. Another group studied collagen biosynthesis, as evidenced by skin biopsy results, in patients before and after receiving high-dose D-penicillamine 1 g/day for at least 6 months.[99] Among the 13 patients studied, a 36% reduction was noted in mean collagen biosynthesis (p<0.025), which correlated well with total dose of D-penicillamine given (r=0.71, p<0.01). This reduction in collagen synthesis may contribute adversely to wound healing. A similar study was designed to evaluate the effects of D-penicillamine treatment with or without long-term corticosteroid use on wound healing in 21 patients with rheumatoid arthritis (mean age 51.1 yrs) who underwent a total of 42 orthopedic surgical operations.[90] The authors observed no difference in the mean time to wound healing in patients taking D-penicillamine compared with patients taking both D-penicillamine and prednisolone (mean B1 SD 19.8 B1 13.1 vs 19.3 B1 10.1 days, respectively). The mean dose of D-penicillamine taken was 750 mg, which is higher than what is commonly used in practice.

Taken together, these studies give little evidence to justify withholding D-penicillamine before surgery because of the potential for disruption of collagen synthesis and delays in wound healing.

Azathioprine. Azathioprine is an antimetabolite that inhibits purine metabolism, synthesis of nucleic acids and proteins, and cellular mitosis. Because of its antimetabolic effect, azathioprine was hypothesized to have adverse effects on wound healing. This hypothesis was tested in four subgroups of rats (9-14 rats in each group).[91] Investigators made abdominal incisions and administered either azathioprine alone 20 mg/kg intraperitoneally, prednisolone alone 0.5 mg/kg intramuscularly, or azathioprine plus prednisolone for 2 days before wounding followed by daily administration from days 3 through 7. The fourth group was the control group. After sacrificing the rats, the abdominal wall was excised and tested for breaking strength with a motor-driven tensiometer. Surprisingly, the breaking strength of tissue from rats that received azathioprine alone was significantly stronger compared with controls (1355 vs 1223 g, p<0.05). However, two attempts to reproduce this finding were unsuccessful. The investigators concluded that although the dose examined in rats was much higher than that commonly used in humans, azathioprine has no inhibitory effect on wound healing in rats.

In a study conducted to evaluate the risk factors for surgical wound complications in patients with rheumatoid arthritis,[82] a total of 204 patients underwent 367 orthopedic surgeries that included reconstructive joint surgery, namely hip and knee replacements. By univariate analysis, the study found that the preoperative use of azathioprine (used 32 times in the 204 patients) was associated with approximately twice the risk of postoperative complications (7% vs 16%, relative risk 2.13, 95% CI 1.04-4.37). However, a multivariate analysis was performed and revealed no association between DMARD use and surgical complications. Note that patients in this study who received azathioprine had a greater frequency of previous DMARD use, which suggests that their disease was more severe and possibly associated with greater morbidity.

A number of studies have evaluated the effects of immunomodulators on intraabdominal septic complications in the perioperative treatment of Crohn's disease ( Table 3 ).[100,101,102,103] One group retrospectively examined the effect of peri-operative treatment with methotrexate, infliximab, azathioprine, or its metabolite 6-mercaptopurine, on early complications after colonic resection, strictureplasty, or both.[100] Patients were included if they received azathioprine or 6-mercaptopurine within 4 weeks of surgery, but no minimum dose was required. Of the 270 patients studied, 102 were receiving either azathioprine or 6-mercaptopurine at the time of surgery and 52 had received infliximab. Patients received azathioprine 0.5-2.7 mg/kg/day (median daily dose 2.0 mg/kg), which corresponds to commonly used dosages in clinical practice.[104,105] Patients receiving any dose of an immunosuppressive agent (azathioprine, 6-mercaptopurine, or methotrexate) were grouped together and found to have no significant increase in septic or total complications by logistic analysis (19% septic complication rate, odds ratio [OR] 1.0, 95% CI 0.5-1.8; 25% total complication rate, OR 1.1, 95% CI 0.6-2.0) after abdominal surgery. The authors concluded that there was no association between postoperative complications, including intraabdominal infections, and the perioperative use of azathioprine or 6-mercaptopurine. Likewise, a univariate regression analysis showed no effect of any infliximab administration on the rate of total or septic complications.

In a similar retrospective study, investigators hypothesized an improved perioperative outcome after surgical intervention and immunomodulation with azathioprine, 6-mercaptopurine, methotrexate, or infliximab.[101] They believed that the suppression of inflammation associated with Crohn's disease by immunomodulators would decrease the frequency of abscess, fistula formation, and anastomotic leak during the first 4 weeks after surgery. Patients were included in the analysis if they underwent a first colon resection with anastomosis or strictureplasty. Subjects (72 patients, median age unknown) in the immunomodulator-receiving group were required to have maintained treatment with immunomodulators or biologic agents for 8 weeks before surgery. Of the 28 patients in the nonimmunomodulator-treated group, 14 were receiving corticosteroids. Analyses showed that 4 (5.6%) of 72 patients receiving immunomodulation developed intraabdominal septic complications, whereas 7 (25%) of 28 patients not receiving immunomodulation developed intraabdominal septic complications (p=0.01, OR 5.68, 95% CI 1.51-21.3). This finding, though congruent with the hypothesis, is contrary to the traditional thought that immunosuppression may delay surgical healing and increase postoperative complications.

Though sparse data are available, results of the above studies indicate that immunosuppressive agents such as azathioprine do not adversely affect the wound healing process in the postoperative period for most patients. These studies, however, did not include patients with rheumatologic disease. Since patients with rheumatoid arthritis are less well studied, we must draw our conclusions from the available literature as well as clinical experience. Thus, it is our opinion that the risk of postoperative complications with azathioprine use is outweighed by its beneficial effects in maintaining disease remission.

Leflunomide. Leflunomide is approved for the treatment of active rheumatoid arthritis.[106] After oral administration, leflunomide is converted to an active metabolite, A77-1726. This metabolite has a half-life of 11 days and is equally excreted in bile and urine.[6] Its long half-life appears to be due to enterohepatic recycling. It has been proposed that the antirheumatic properties of leflunomide are propagated as the active metabolite inhibits the mitochondrial enzyme dihydroorotate dehydrogenase, which causes a depletion of pyrimidine nucleotides. Ultimately, G1 phase cell cycle arrest occurs, leading to the halting of stimulated lymphocyte proliferation.[107,108] In addition, leflunomide may exert an immuno-suppressive action by elevating transforming growth factor-β levels[109] and an antiinflammatory action through the inhibition of COX-2.[110] Leflunomide inhibits adherence of leukocytes to vascular endothelial cells, which may have a negative effect on the process of wound healing.[111]

Leflunomide therapy may also lead to myelosuppression. A recently published case series reported the development of pancytopenia in 18 patients taking leflunomide.[6] Approximately 80% of patients who developed pancytopenia were taking methotrexate concomitantly. In most cases, patients required hospital admission and were treated with supportive care and withdrawal of immunosuppressive therapy. The development of severe neutropenia in patients taking leflunomide with other immunosuppressants may increase the risk for infection; consequently, patients undergoing surgery while taking leflunomide with or without other immuno-modulating drugs may be at an additional risk for postoperative complications.[106] Despite the above mechanisms, studies are lacking to show the effect of leflunomide or A77-1726 on wound healing.

If severe adverse effects from leflunomide that preclude surgery are present, A77-1726 elimination can be expedited with the administration of activated charcoal or cholestyramine by using a "washout procedure." The procedure involves the administration of cholestyramine 8 g 3 times/day orally or activated charcoal 50 g/day orally over 11 days.[112]

Other DMARDs such as gold sodium thiomalate, sulfasalazine, and hydroxychloroquine have limited data on their effects on the wound healing process. One retrospective study was conducted to determine the risk factors for postoperative wound complications in patients with rheumatoid arthritis who were taking DMARDs.[82] Treatment with oral and parenteral gold salts (used 87 times), methotrexate (66), azathioprine (32), hydroxychloroquine (26), and D-penicillamine (25) were allowed until immediately before surgery, most of which were total knee replacements. All patients in the study received at least one preoperative antibiotic. Univariate comparisons revealed no statistically significant risk of postoperative complications with the use of gold salts (RR 0.87, 95% CI 0.47-1.6), hydroxychloroquine (RR 1.63, 95% CI 0.66-4.02), or sulfasalazine (RR 1.55, 95% CI 0.32-7.38) in orthopedic surgery. Hydroxychloroquine appears to have a lower toxicity profile, and we have no data suggesting that it should be held perioperatively. Although this retrospective study found no increased risk of postoperative complications associated with the use of gold salts, this therapy is rarely used in the management of rheumatoid arthritis due to its adverse-effect profile.

Etanercept, infliximab, and adalimumab exert their pharmacologic action by blocking TNF-α cell surface receptors (also circulating TNF-α) and have found a place in the treatment of active rheumatoid arthritis, Crohn's disease, psoriasis, psoriatic arthritis, juvenile arthritis, and ankylosing spondylitis. They can be administered as intravenous or subcutaneous injections, and due to their long half-lives, they can be given at extended dosing intervals of between 1 and 8 weeks. In addition to inhibiting angiogenesis and fibroplasia, this pharmacologic class has the potential to influence wound healing by decreasing the expression of adhesion molecules, such as E-selectin and intercellular adhesion molecule-1, which are needed for leukocyte migration and recruitment.[113] Thus, immune system impairment may occur.

Tumor necrosis factor-α functions to stimulate angiogenesis and fibroblast proliferation, as well as to increase collagenase and prostaglandin synthesis.[114] In addition, the TNF-α receptor has been shown to be necessary for an effective immune response to mycobacteria in animal models.[115] This is corroborated by reports of active tuberculosis in patients receiving etanercept or infliximab.[3,4,5] Two major trials have evaluated the efficacy and safety of the newest TNF-α antagonist, adalimumab, with a few cases of tuberculosis reported in the literature.[116,117]

In addition to the above, reports of histoplasmosis, coccidiomycosis, listeriosis, pneumocystosis, and other fungal, bacterial and mycobacterial infections have been observed in patients receiving infliximab, etanercept, and adalimumab.[118] Infectious complications prompted the manufacturers of infliximab and etanercept to include black-boxed warnings on their product labeling about the risk for serious infection and sepsis leading to death in patients who are being treated. In addition, the manufacturer of etanercept advises the use of caution when prescribing the agent in patients who have either a history of recurring infections or an underlying condition such as advanced diabetes, which may be a predisposing factor to infection.[113]

One group investigated the effects of TNF-α inhibition with etanercept or infliximab on infectious or healing complications after orthopedic foot and anklesurgery.[102] Thirty-one patients with rheumatoid arthritis were enrolled in this prospective study. Group 1 (mean age 50 yrs, range 40-66 yrs) consisted of 16 patients receiving active treatment with either etanercept twice/week by subcutaneous injection or infliximab intravenous infusion every 8 weeks (doses not specified). The treatment group was compared with patients with rheumatoid arthritis who were not being treated with TNF-α inhibitors (group 2, mean age 60 yrs, range 41-73 yrs). No differences were noted between the two groups with respect to usage of NSAIDs, methotrexate, leflunomide, or corticosteroids (doses were not specified). Patients taking etanercept received their last preoperative dose 1-10 days before surgery (mean 2.6 days), whereas patients taking infliximab received their last preoperative dose 2-45 days before surgery (mean 20.2 days). Both groups resumed their respective therapies immediately after surgery according to their previous dosing schedule. All patients were given perioperative antibiotic prophylaxis for 72 hours. Despite a higher percentage of smokers in group 1 (37.5% vs 0.0%, p=0.02), patients in group 2 experienced a higher overall complication rate (rate not given, p=0.033) at a mean follow-up of 9.7 months. Investigators concluded that TNF-α inhibitors could be safely administered in the perioperative period without increasing the risk of infectious or healing complications.

Product labeling for infliximab, etanercept, and adalimumab specify the potential risk for hematologic events, including leukopenia, neutropenia, thrombocytopenia, pancytopenia, and aplastic anemia.[113,118] Discontinuation of therapy should be considered for all patients who develop significant hematologic abnormalities. Despite the findings in the above-mentioned study,[102] it seems reasonable to assume that the development of an infection by opportunistic pathogens (as a result of treatment with a TNF-α inhibitor) places the patient at increased risk for postoperative complications.

Additional prospective studies are needed to examine the frequency and relationship of postoperative complications with the use of etanercept, infliximab, or adalimumab among patients with autoimmune disorders. Until these studies are performed, we can only examine the data regarding the use of TNF-α antagonists in other patient populations. One group conducted a retrospective study to evaluate the effects of infliximab on healing rates of fistulizing anorectal Crohn's disease.[103] Twenty-nine patients (mean age 31 yrs) were included who had multiple, longstanding, or previous anorectal fistulas that failed surgical and/or standard medical treatment. Infliximab induction consisted of 5-mg/kg doses at 0, 2, and 6 weeks, with additional infusions titrated as needed for clinical response. A complete or partial response to therapy was seen in 86% of cases. This study has several limitations that make it difficult to draw definitive conclusions on the effect of infliximab in wound healing. However, this small study did not show worsening of healing rates in patients with anorectal fistulas.

In a similar study (discussed in the azathioprine section),[100] the authors examined the safety of corticosteroids, azathioprine, 6-mercaptopurine, methotrexate, or infliximab after colonic resection, strictureplasty, or both in patients with Crohn's disease. The median patient age was 40 years, and the most common indication for surgery was colonic stricture. This retrospective analysis included 52 patients who had received infliximab 5 mg/kg within 8 weeks before surgery or 4 weeks after surgery, with or without other immunosuppressives (proportion of individuals receiving each agent was not specified). Patients receiving any infliximab had a 17% rate of septic complications (OR 0.9, 95% CI 0.4-1.9) and a 23% rate of any complication (OR 1.0, 95% CI 0.5-2.0) within the first 30 days after surgery. Univariate regression analyses showed no effect of concomitant corticosteroid or immunosuppressive administration on the development of postoperative complications. A multiple regression analysis could not be performed because of the low occurrence of postoperative complications. This led the authors to believe that there was no significant association between infliximab therapy and postoperative complications, and they concluded that infliximab can be safely administered perioperatively to patients undergoing abdominal surgery for Crohn's disease.

In the treatment of perianal and rectovaginal fistulas, colonic stricture or perforation, or refractory Crohn's disease, infliximab infusion plus surgical intervention does not significantly compromise postoperative wound healing. There is, however, a documented risk of developing cytopenia and infection with the use of infliximab or etanercept. Physicians should weigh the risk of immunosuppression and/or infection against the benefit of primary disease management when deciding whether to discontinue treatment in the perioperative period. Patient-specific characteristics, such as concomitant disease states and history of hematologic abnormalities, should be considered when making treatment decisions.

Anakinra is the first in a novel class of drugs that blocks the interleukin (IL)-1 receptor.[119] The drug is indicated for reducing the signs and symptoms and slowing the progression of structural damage in patients with moderately to severely active rheumatoid arthritis who are at least 18 years of age and have failed treatment with one or more DMARDs. By competitively blocking the IL-1 type I receptor, anakinra reduces the inflammatory disease response. To our knowledge no data are available that evaluate its effects on surgical outcomes. A recent post hoc analysis investigated the safety of anakinra when used at standard daily doses in patients with active rheumatoid arthritis and other comorbid conditions such as coronary artery disease, chronic obstructive pulmonary disease, diabetes mellitus, or renal impairment.[120] The authors found injection-site reactions to be the most common adverse event, and the rate of serious adverse events was comparable between patients receiving anakinra (8.6%) and those receiving placebo (8.7%). There was, however, an increase in the rate of serious infectious events in patients receiving anakinra compared with those receiving placebo (2.5% vs 0.0%). Most of these events were classified as pneumonia, but not due to opportunistic infections such as tuberculosis or histoplasmosis. Because of limited data regarding the use of anakinra in the perioperative period, practitioners should again take into account patient-specific variables including risk of infection when making decisions regarding its administration before and after surgery. In accordance, when the risk of infection outweighs the potential benefits of anakinra, practitioners should discontinue usage until the patient's risk of infection has diminished.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: